DETERMINANTS OF FOREIGN DIRECT INVESTMENT: GLOBALIZATION INDUCED CHANGES AND THE ROLE OF FDI POLICIES * 1. Introduction John H. Dunning Emeritus Professor of International Business Rutgers and Reading Universities One of the most remarkable features of the last decade has been the rapidly growing transnationalisation of most economies in the world. As set out in Exhibit 1, and expressed as a percentage of the global gross domestic product (gdp), the share of the combined inward and outward foreign direct investment (fdi) stocks rose from 19.2% in 1990 to 34.0% in 1999 and an estimated 38.2% in 2000. However in 2001, mainly due to an (estimated) fall of world fdi flows by 40%, this ratio is likely only to have increased marginally - to around 39.0%. Yet even from these figures, it can be seen that fdi flows still remain one of the most dynamic constituents of the global economy; and it is expected that, in 2002, as in 2001, they will considerably outpace the growth of both trade and world gdp. In this paper, we shall be primarily concerned with the impact of recent economic and political events, and the likely course of future events, on the geography of fdi; and in particular, on the role of fdi policies in influencing the location of the value added activities of multinational enterprises (MNEs). To what extent is the current slow down in economic growth, the realignment of some exchange rates, and the aftermath of September 11th, 2001, causing international firms to retrench, or geographically realign, their foreign production; or to service their global markets by way of exports, or by engaging in more sub-contracting with foreign suppliers? Or will the burgeoning of regional integration schemes; the cautious optimism arising from the WTO discussions in Doya; the impact of e-commerce; the need to continue to * Part of this paper was originally published as an article by the present author in the Economist Intelligence Unit's World Investment Prospects, 2002, London 2002.
tap into foreign sources of knowledge and other created assets; and the promising opportunities for foreign investors now being opened up by the larger developing countries - notably China and India, - counteract any short-term cross-border risk averse strategies. EXHIBIT 1 COMBINED INWARD AND OUTWARD FDI STOCKS AS A PERCENTAGE OF GROSS DOMESTIC PRODUCT 1985-2001 (%) 1985 1990 1995 1997 1998 1999 2000 2001 World 14.2 19.2 20.5 23.6 27.8 34.0 38.2 39.0 Developed Countries Developing Countries Central and Eastern Europe 13.6 18.2 21.0 24.4 28.5 33.5 38.4 39.1 15.7 16.0 20.4 22.4 26.7 38.1 38.2 38.9 neg 1.8 6.0 9.5 13.8 15.1 15.2 15.3 Source: (i) (ii) 1985-1999 UNCTAD World Investment Report (various issues) 2000 and 2001 Author's estimates. Factors influencing recent trends in FDI The data in Exhibit 2 suggest that the later 1990s saw a redirection in global fdi inflows to the developed regions of the world - and especially Western Europe. There were three main reasons for this. First, there was a huge cross-border merger and acquisition (M&A) boom, which overwhelmingly involved MNEs from one part of the Triad buying firms in another part of the Triad. Something approaching 70% of all new fdi was accounted for by M&As between 1995 and 2000. 1 Most of these were geared towards augmenting the technological, managerial or entrepreneurial assets of the acquiring firms or accessing new markets. In 2000, no less than 73% of all cross-border sales and mergers were within the services sectors (and four fifths of 2
these were in transport and communications, finance and business services). In manufacturing, deals in the electrical and electronic equipment and food, drink and tobacco sectors headed the list. The slow down in economic growth, particularly among the Triad countries, dramatically reduced the number of cross-border M&As in 2001. Even so, in the first nine months of that year, although only about one-half of the value of the same period of the previous year (at $400 billion), the cross border M&A sales were still higher than those for the average of the preceding six years. Again, most of these involved European or trans-atlantic corporations. Second, the growth of regional integration schemes, especially the completion of the internal market in Europe and the maturation of NAFTA has also had an important effect on the geography of MNE activity. Exhibit 3 shows a steadily rising share of fdi inflows, accounted for these two latter regions in the latter half of the 1990s. Preliminary data for 2001 suggest a considerable decline of that share to around 63%. Nevertheless, research, both in the UK and the US, clearly shows that as a direct result of reduced trade and investment barriers in the 1980s and early 1990s, both intra- and extra-eu and NAFTA cross border merger activity increased; and in the latter part of 2002 and 2003 one might reasonably expect these shares to gain ground again. Third, the slow down in the economic growth of mainland China, and the Eastern Asian economic crises of the mid-1990s, considerably reduced the earlier attractiveness of the region to foreign investors: indeed, between 1996 and 1999, its share of new South and East Asian fdi inflows, as a percentage of world inflows, fell from 23.2% to 9.0%. Since then, mainly due to the restructuring of Korean economic policy, and the depreciation of the won, together with the renewed optimism by foreign investors in the Chinese (and particularly the Hong Kong) 1 We accept, of course, that there is no automatic one-to-one coincidence between fdi flows and cross-border M&A deals. For further details see UNCTAD (2000). 3
EXHIBIT 2 SHARE OF FDI INFLOWS ACCOUNTED FOR BY MAJOR REGIONS OF THE WORLD 1995/2001 (%) 1995 1998 1999 2000 2001 (est) 1 DEVELOPED COUNTRIES (of which North America, Western Europe, Japan). 62.0 (20.5) (35.4) (0.1) 69.3 (28.4) (39.5) (0.5) 77.2 (29.8) (45.1) (1.2) 79.1 (27.1) (49.8) (0.6) 67.1 (22.0) (44.6) (0.5) 2. DEVELOPING COUNTRIES (of which Africa Latin America and the Caribbean Asia and the Pacific) 1 3.2 (1.4) (9.8) (22.9) 27.1 (1.2) (12.0) (13.8) 20.7 (0.8) (10.2) (9.3) 18.9 (0.6) (6.8) (11.3) 29.6 (1.3) (10.5) (16.4) 3. CENTRAL AND EASTERN EUROPE 2 4.3 3.2 2.3 2.1 3.6 WORLD 100.0 100.0 100.0 100.0 100.0 1. Of which China, including Hong Kong (12.7) (8.5) (6.4) (8.3) (12.5) 2. Including countries in former Yugoslavia. Source: 1995/2000 UNCTAD World Investment Report (various issues). 2001 UNCTAD and Author's estimates. 4
EXHIBIT 3: SHARES OF INWARD FDI DIRECTED TO TWO AREAS OF REGIONAL ECONOMIC INTEGRATION 1990-2001 (%). 1989/94 1 1995 1998 1999 2000 2001 (est) European Union 38.3 34.3 37.7 43.5 48.6 38.0 NAFTA 27.4 23.4 30.1 30.9 28.1 24.5 65.7 57.7 67.8 74.4 76.7 62.5 Source: 1989-2000 UNCTAD World Investment Report (various issues) 2001 Author's estimate. economy, it recovered to 10.8% in 2000. 2 In 2001, it is estimated to have risen even more sharply to 15.2%. What, however, of the specific affect of September 11th? Most certainly the immediate impact has been to accentuate the economic downturn of the previous nine months, and to create a higher level of corporate uncertainty. In the medium term, the new global economic environment is likely to have two major affects on the geography of fdi. First, by placing a higher premium on environmental risk, it is likely to steer MNE activity towards locations perceived to be friendly towards their home country regimes; and be least subject to terrorist attacks or political instability. Second, it may well add to the competitive pressures in many industries forcing companies to enhance their cost efficiency. Faced by more price driven competition, some MNEs may choose to relocate some of their production facilities in low (real) cost developing countries. At the same time, due to enhanced economic vulnerability, other MNEs may choose to reduce their foreign equity stakes, particularly in advanced technology 2 The dramatic increase in fdi flows into Hong Kong is put down to (a) a general improvement in the local business environment (b) China's imminent succession to the WTO (c) a $23 billion merger deal involving China M obile, but financed from new shares issued in the British Virgin Islands and (d) the emergence of Hong Kong as a funding hub for new business in the region (UNCTAD) 2001, p.25.) 5
sectors, - and conclude more strategic alliances and sub-contracting arrangements with foreign firms.. The changing locational determinants of fdi. Most empirical studies on the locational determinants of fdi recognise that these are strongly dependent on (a) the motivation for the fdi (cf. natural resource seeking with market seeking, efficiency seeking and asset augmenting objectives.) (b) the economic and business environment of host, or potential host, countries, and the fdirelated policies pursued by their governments. (c) the mode of entry or expansion of the fdi (cf. greenfield fdi with mergers and acquisitions (M&As)). A scheme identifying these locational determinants is set out in Exhibit 4. Over recent years much of scholarly research and a plethora of business surveys on the host country determinants of fdi has been concerned with evaluating the significance of such variables set out in this exhibit; 3 and most noticeably that of policy, tax and investment incentives, labor costs, market size and the policy framework for fdi. 4 However, our focus in this paper is on the changes in the significance of some of these variables over the last decade or so. To do this, we think it useful to look at the four types of FDI set out in Exhibit 5; and of how both the principal economic determinants and the response of host country governments to these determinants, have affected, and are affecting the locational 3 The more important of these are identified, and in some cases, reviewed in UNCTAD (1998). 4 In this connection see a thoughtful paper on locational competition and the appropriate policy instruments to attract the appropriate fdi - or least (real) cost (Siebert 2000). 6
strategies of mobile investors. 5 While the contents of the Exhibit are largely self-explanatory, we would emphasise a couple of points related to the first two kinds of MNE activity. (1) Developed - developing country fdi. Besides the traditional market resource seeking fdi, recent intra-developed country MNE activity has been dominated by two kinds, each of which responds differently to host country determinants. The first, as we have already mentioned, is by way of the M&As which, in 2000, accounted for 95% of all cross border M&A purchases and 92.4% of all such sales. These are essentially motivated by the expectation that the acquired firms help the acquiring firms (a) to upgrade their competitive advantage and/or (b) gain access to new markets or supply capabilities. In this respect, the host country's business environment, and government regulations with respect to competition policy innovatory and cross-border M&As, is proving to be an increasingly influential locational determinant (in Europe, cf the attitude of the German authorities towards cross-border M&As compared with those of the UK). The second kind of intra-developed country fdi is horizontal efficiency-seeking fdi, which is being encouraged inter alia by market liberalization and regional integration schemes. Here the critical locational determinants are increasingly comprising those identified under C and III rather than A,B and D and I in Exhibit 4. (2) Developed - developing country (FDI). Again, it is worth identifying two main forms of fdi. The first is the traditional market seeking or resource seeking fdi, which, excluding taxhaven related fdi, still accounts for the majority of North and South fdi (probably around 70%) - and particularly for that in the larger countries like China, India and Brazil. 6 Here the critical variables are those identified by A B and 1 and 2 in Exhibit 4. Such fdi tends to be fairly location bound - though the policy and business facilitating framework may be critical in determining whether an fdi takes place at all. 5 Mobile investors may include both foreign owned and domestic owned MNEs. 6 This figure depends on whether or not (for statistical purposes) one treats Hong Kong as part of China. If one does not, then the 70% figure drops to around 50%. 7
The second type of fdi is vertical efficiency seeking fdi in which foreign companies seek to produce intermediate and/or final products in the cheapest (real) cost locations primarily for exports to third markets. Here the variables identified by C and II III in Exhibit 1 are the critical ones - given a satisfactory (general) policy framework for fdi (and indeed domestic investment). Currently, such investment is largely concentrated in South, East and South-Asia, and Mexico. Impact of the value chain on the geography of FDI So far in this paper, we have been considering the geography of all foreign activities of MNEs financed by fdi. However, it is well known that not only is this geography of such activities and the policies of host governments towards them, likely to vary between industrial sectors, but also, according to the raison d'etre for those activities, and the particular parts of the value chain which MNEs choose to locate outside their home countries. The fact that a good deal of intra-triad fdi is now being made to protect or augment the existing competitive advantages of the investing firms rather than exploiting such advantages, suggests a very different geography of MNE activity, both between developed countries and within them. By contrast, by far the greater part of fdi in developing countries (notably in the larger and faster growing economies) is directed to accessing local natural resources and/or national (or regional) markets. The future locational strategy of MNEs and the impact of the fdi related policies of host governments is then likely to rest on the relative significance and growth of the two types of fdi. This, in turn, will reflect the opportunities offered by local markets and supply capabilities on the one hand, and the need to tap into the global panorama of resources and capabilities - particularly that of all kinds of knowledge - on the other. On the question of which functions MNEs are most likely to locate outside their home countries, the literature is fairly clear. In a survey conducted by the present author at Rutgers University five years ago, it was found that, although 40% of the global assets of 150 of the world's leading industrial companies were located outside their home countries, only 23.3% of 8
their research and development (R&D) activities was so sited. The tendency of MNEs to concentrate their innovatory activities in their home countries - usually for good economic reasons - is also confirmed by data on the extent to which the proportion of patents registered by the world's largest firms are attributable to research undertaken by their foreign affiliates. However, as John Cantwell of Reading University has shown, this proportion has gradually risen over the years to an average of around 11.5% in the mid 1990s; and is highest in the case of European and lowest in the case of US and Japanese firms (Cantwell and Janne, (2000)).. Furthermore, in 2000, whereas 68.5% of the stock of fdi by MNEs was located in developed countries, (the proportion for 2001 was nearer 65%), between 80 and 85% of their innovatory activities were so located. Though there is an increasing tendency for MNEs to site some of their research and development (mostly development) activities in the larger or higher income developing countries (notably Korea, India, China, Singapore and Brazil), the great bulk continue to be performed within the Triad; and, it is a fact that a growing number of the more knowledge or information intensive MNEs are desiring, and obtaining, an active R & D presence in each of the main Triad regions (Kuemmerle 1999). Research has also shown that such a presence is offering handsome benefits to the investing companies, both in the access it allows to new sources of technological, organisational and marketing expertise, and to foreign innovatory systems, and by the R & D actually undertaken by the local affiliates of the MNEs. It is also causing host governments to reappraise their technology policies in the light of the increasing presence of innovatory activities by MNEs. 7 Other factors affecting the location of FDI We would make three other brief observations about the recent and likely future locational pattern of MNE activity. The first concerns the likely impact of the internet - and 9
more particularly business to business (B to B) electronic commerce. While there are a few examples of tangible goods (which need to be produced in a particular location), being replaced by services downloaded on the computer, the major effect of the internet is currently on the mode of delivery of the goods and services in question (Dunning and Wymbs, 2001). For if nothing else, the internet is a market facilitating instrument, as it reduces intra-firm information asymmetries and spatial transaction costs, both along and across value chains. In so doing, it has enormously widened the locational options open to firms in seeking out their suppliers, customers and possible collaborators. This is particularly so where reasonably standardised commodities are being produced and sold; and where the necessary information and knowledge, e.g. in the form of designs, blueprints and specifications, can be transferred and utilised across national boundaries without difficulty. Supply chains in such sectors as autos, garments, and electronic goods are not only becoming dis-internalised by firms, but are becoming more spatially dispersed. The internet is, indeed, a major force for globalisation and widening the locational choice of MNEs (Zaheer and Manrakhan 2001). Moreover, in the case of the more labour intensive activities, it has considerably favored the economies of lower and middle income developing countries that pursue the right policies towards, or in the light of, inbound fdi. At the same time, the advent of E-commerce is having much less impact on the geography of the higher value and more idiosyncratic activities of firms. As these tend to involve market failures or distortions which the internet is unable to correct. This being so, we foresee only a modest geographical dispersion of the innovatory activities of MNEs in the years to come. We believe these are likely to remain concentrated in a relatively few locations where the resource capabilities and infrastructure are congenial to such activity. 7 These issues are explored by various authors in Archibugi, Howells and Michie (1999). 10
The second (and related) new element affecting the location of international business is the trend towards a sub-national agglomeration of related activities by firms. 8 Sometimes this clustering is among firms producing similar products; sometimes between firms and their suppliers and customers. While the principle of comparative advantage has long since helped explain the composition of production between countries, it is now recognised that, in the larger economies of the world at least, the principle is no less relevant in explaining the spatial distribution of some kinds of activities within countries. Moreover, the advantages of spatial agglomeration (e.g. access to common learning facilities, information, infrastructure, factor and product markets) are no less applicable to services as to manufacturing activities; and can be gained in both developing and developed countries. And it seems likely that MNEs will be increasingly drawn to such sub-national groupings in the next decade or more. Finally, we would mention the growing trend towards the setting up, diversification and expansion of the regional headquarters (RHQs) of MNEs. This trend is again evident both in the case of developed and developing countries; and is lending support to Alan Rugman's assertion that MNE activity is becoming increasingly regionalised rather than globalised (Rugman (2000)). 9 Certainly, it is being encouraged by the growing number of regional integration schemes; and the extent to which MNEs engage in intra-regional cf inter-regional trade. 10 Such RHQs are, by no means, confined to the advanced industrialised countries or to firms engaged in manufacturing industry. Both Hong Kong and Singapore house the RHQs of Asian based trade, shipping and financial and business service firms, most of the value added activities of which are conducted within the region (Enright, 2000). Once again, we see, in the RHQ phenomena, a tendency both towards more geographical diversification, but also to an increased specialisation 8 There is a large literature developing on this subject. For a recent review of this literature see Dunning (2000). 9 Alan Rugman The End of Globalisation, London, Random House Business Books. 10 For example, in the late 1990s some 75% of the exports of US subsidiaries in EC countries are intra-ec; and 80% of those in Mexico and Canada are intra NAFTA (US Department of Commerce data). 11
within regions to take advantage of the economies of scale, access to high quality skills and services, and a strategic location. Conclusions To conclude: the evidence strongly suggests that there are both short and long term forces compelling MNEs to rethink their locational strategies, and for national and sub-national governments to reappraise their micro-management policies designed to attract and retain the 'right' kind of fdi to their borders. As to the former, it is becoming increasingly evident that spatially related decisions, notably in respect of what and how much to sub-contract, what and how much to produce, and to whom and where to sell end products, is becoming a more important component of competitive advantage. While it remains important that corporations internally generate their core competences, it is also essential that they efficiently use these - many of which are geographically mobile - jointly with other resources and capabilities in different parts of the world which are spatially less mobile. However, as we showed earlier, the locational strategies actually chosen by firms are likely to be highly contextual; and to vary according to industry specific characteristics, the motives for fdi, and the functions being performed by MNE subsidiaries. In this paper, we have also suggested that, quite apart from the impact of the current economic slow down and the events of September 11th, the internet, the widening scope of the knowledge based economy, and regional integration schemes are also affecting the geography of fdi. As to the role of governments as they seek to attract MNE activity, we believe they need to recognise that the location specific advantages sought by mobile investors are changing. While in some countries, e.g. the larger developing countries, such traditional economic variables, e.g. the availability of cheap labour, natural resources and market size, remain important, in others, e.g. the more advanced industrialised countries, MNEs are increasingly seeking complementary knowledge 12
intensive resources and capabilities, a supportive and transparent commercial, legal communications infrastructure, and a gamut of government policies favourable to globalisation, innovation and entrepreneurship. Field surveys of the rankings of various countries by business executives compiled and published by the EIU in the 2000 and 2001 editions of the World Investment Prospects (E.I.U. 2002) show that business executives are increasingly ranking the political stability, quality of infrastructure and government policies towards private enterprise and competition, along with the macro economic environment, as the critical variables likely to affect the future geography of fdi in the early years of the 21st century. To these, the propensity to terrorist attacks or other destabilizing events might be added. These results, which are broadly confirmed by other studies on the investment intentions of firms, are useful pointers to governments wishing to attract such investment. They also suggest that these same governments need to give constant attention to the upgrading and reconfiguring of their own unique location bound advantages; (both actual and potential); and to target the kind of fdi which might help them best accomplish this objective. Rutgers and Reading Universities 13
REFERENCES Archibugi, D. Howells, J. and Michie J.(1999), Innovation Policy on a Global Economy Cambridge University Press. Cantwell, J. and Janne, O. (2000) The role of multinational corporations and national states in the globalization of innovatory capacity: the European perspective. Technology Analysis and Strategic Management, 12(2), 243-262. Dunning, J.H. and Wymbs, C. (2001) The challenge of electronic commerce for international business theory, International Journal of the Economics of Business, 8(2), 273-302. Dunning, J.H. (ed.) (2000), Regions, Globalization and the Knowledge Based Economy, Oxford, Oxford University Press. Economist Intelligence Unit (2001) World Investment Prospects, London E.I.U. Economist Intelligence Unit (2002) World Investment Prospects, London, E.I.U. Enright, M.J. (2000) Globalization, regionalization and the knowledge based economy in Hong Kong in Dunning, J.H. (ed.) (2000), Regions, Globalization and the Knowledge based Economy, Oxford, Oxford University Press, 381-406. Kuemmerle, W. (1999) The drivers of foreign direct investment into research and development: an empirical investigation, Journal of International Business Studies, 30(1), 1-24. Rugman, A.M. (2000) The End of Globalization, London, Random House Books. Siebert, H. (2000) The Paradigm of Locational Competition. Kiel: Institut fur weltwirtschaft Kiel Discussion Papers, No.367, August. UNCTAD (2001) World Investment Report: Promoting Linkages, New York and Geneva, U.N. UNCTAD (2000) World Investment Report: Cross Border Mergers and Acquisitions and Development. New York and Geneva. UNCTAD (1998) World Investment Report: Trends and Determinants, New York, and Geneva, U.N. Zaheer, S. and Manrakhan, S. (2001) Concentration and dispersion in global industries: remote electronic access and the location of economic activities, Journal of International Business Studies, 32(4), 667-86. 14
15
EXHIBIT 4 HOST COUNTRY DETERMINANTS OF FDI HOST COUNTRY DETERMINANTS I. Policy framework for FDI economic, political and social stability rules regarding entry and operations standards of treatment of foreign affiliates policies on functioning and structure of markets (especially competition and M&A policies). international agreements on FDI privatization policy trade policy (tariffs and NTBs) and coherence of FDI and trade policies tax policy industrial/regional policy II Economic determinants III Business facilitation investment promotion schemes including image-building and investment-generating activities and investment-facilitation services) investment incentives reduced hassle costs related to corruption, bureaucratic inefficiency. social amenities (bilingual schools, quality of life, etc.) pre- and post-investment services (e.g. one stop shopping. protection of property rights good infrastructure and support services e.g. banking, legal, accountancy services economic morality. Types of FDI classified* Principal Economic determinants in host by motives of TNCs countries A.Market-seeking market size and per capita income market growth access to regional and global markets country specific consumer preference structure of markets B. Resource seeking land and building costs/rents and rates raw materials, components, parts. low-cost unskilled labour skilled labour C. Efficiency-seeking cost of resources and assets listed under B adjusted for productivity for labour inputs. other input costs, e.g. transport and communication costs to/from and within host economy and costs of other intermediate products. membership of a regional integration agreement conducive to promoting a more cost-effective and product upgrading inter-country division of labour. D. Asset seeking technological, managerial relational and other created assets be they those embodied in individuals, firms or clusters of firms. physical infrastructure (ports, roads, power, telecommunication) macro-innovatory, entrepreneurial educational capacity/environment. *Each of these in turn may be classified by entry mode (e.g. greenfield c.f. M&A), by degree of foreign ownership (100% or joint venture); and by those parts of the value chain of a foreign investor s operations being considered. Source: Author s adaptation of Table IV.1 (p.91) of UNCTAD 2001.
EXHIBIT 5 CHANGING LOCATIONAL VARIABLES AFFECTING FDI 1970/80-1990/2000 (a) Motives/Strategies of Firms 1970/1980 1990/2000 (1) Developed-developed?? Mainly market/horizontal efficiency seeking fdi?? More assets augmenting and (horizontal) efficiency seeking fdi?? Mixture of greenfield fdi (or expansion of same) and M&As?? More M&As/strategic Alliances?? Integrated MNE operations (b) Host Country Determinants?? Multi-domestic MNE operations?? Emphasis on business facilitating variables?? Predominantly FDI policy and economic?? Availability of creative assets determinants affecting market -efficiency seeking fdi?? Agglomerative economies (a) Motives/Strategies of Firms 1970/1980 1990/2000???? (2) Developed-Developing?? Mainly market resource seeking?? More (vertical) efficiency seeking fdi and subcontracting?? Greenfield/joint ventures?? Multi domestic MNE operations?? Predominants fdi policy and economic determinants; especially regulation of incentives towards fdi. (b) Host Country Determinants?? Emphasis switched to using fdi to upgrade competitive advantage?? More attention given to economic policies and business facilitation (a) Motives/Strategies of Firms (3) Developing-developed 1970/1980 1990/2000?? Very little fdi?? As for (1) above. (b) Host Country determinants (a) Motives/Strategies of Firms (4) Developing/Developing 1970/1980 1990/2000?? Almost entirely market/resource seeking?? As for 1970s/1980s, but an increasing amount of efficiency seeking and some asset augmenting fdi.?? Multidomestic MNE operations (b) Host Country Determinants?? As for (2) above.?? As for (2) above